12 



SCIENCE. 



[N. S. Vol. VIII. No. 183. 



attends the comparison of differently colored 

 lights ; each found it hard to make readings 

 agreeing with the other, or even to verify 

 the other's readings. With the flicker in- 

 strument these difliculties disappeared. 



As Professor Eood says, there is little or 

 no more difliculty or fatigue in the use of 

 this instrument than in other optical ob- 

 servations, the approximate setting being 

 quickly made, and the flicker near that 

 point being so slight as not to disturb the 

 eye, yet being distinct enough to require no 

 especially strained attention, unless the 

 illumination is too feeble. I may call atten- 

 tion here once more to the statement in my 

 previous j)aper, that the method vised by 

 Professor Eood, of keeping the photometer 

 and one lamp stationary, moving the other, 

 has certain notable advantages in the use 

 of this instrument, especially in the com- 

 parison of a number of lights difi"ering 

 greatly in color and in brightness, in that 

 the standard light can be kept at a fixed 

 distance from the photometer, and all the 

 measurements made at the same absolute 

 brightness, insuring about the same quality 

 of flicker, and about the same degree of 

 sensitiveness throughout. 



The arrangement, however, is obviously 

 not so sensitive as that in which both lights 

 are stationary and the photometer itself is 

 moved, and probably the latter method is 

 generally preferable when the lights are 

 sufficiently bright. It takes a little longer 

 to make a setting with the flicker pho- 

 tometer than with others, and for this 

 reason it is not easy to use upon an arc 

 light, where the brightness is continually 

 changing. This difficulty, of course, varies 

 greatly with the character of the arc. 



The essence of the flicker photometer 

 lies in this, that the two fields to be com- 

 pared are presented to the eye alternately, 

 in rapid succession, and not as in other in- 

 struments simultaneously, and side by side. 

 There is no attempt to compare the two 



lights. One simply notes the disappearance 

 of a certain recurrent sensation. It is evi- 

 dently necessary that the two fields com- 

 pared should fill exactly the same portion 

 of the whole visual field, and it is desirable 

 that the time taken in transition from one 

 field to the other should be as small a por- 

 tion as possible of the whole time of obser- 

 vation, in order that the retinal shock, to 

 which the ' flicker ' sensation is due, may 

 be sharp and sudden. I am inclined to 

 think that the latter condition is more 

 easily and completely met by a rotating 

 than by an oscillating apparatus. 



Professor Rood's form of the instrument 

 possesses some advantages over the rotating 

 disk. The faces of the large prism are 

 rigid, and the same portion of each of them 

 is always used. A rotating disc is thin, 

 easily bent or distorted, and there is a pos- 

 sibility of difference in the character of the 

 surface of different parts, which, since the 

 whole of it passes in review before the eye, 

 might cause a flicker between different por- 

 tions of the disk itself. A flat disc of zinc, 

 however, to which is firmly glued a sheet 

 of white paper, gives but little trouble, and 

 the line separating the two fields is ren- 

 dered almost imperceptible by filing thin 

 that edge of the disk which passes in front 

 of the ej'epiece. Again, the method of 

 construction of the disk photometer makes 

 it impossible that the two lights to be com- 

 pared should be in the same line parallel to 

 the photometer bar, an arrangement which 

 becomes inconvenient if it is desired to 

 change from this to another photometric 

 apparatus. This difliculty is avoided in the 

 oscillatory apparatus of Professor Rood. 



I may mention here a form of the flicker 

 photometer (described at the Detroit meet- 

 ing of the American Association for the 

 Advancement of Science, but not elsewhere 

 published) which avoids some of the diffi- 

 culties of the disk form and is more con- 

 venient for ordinary photometric purposes. 



